M2M & BigData giving life to the Connected Society – Part 3(5) – Retail

mall

When it comes to bigdata, retail is an industry which is quite mature compared to the ones mentioned in my previous articles, especially if you start looking at companies like Wallmart. The subject of bigdata is already well covered but I hope I can give an additional twist on the current developments by taking M2M into the picture.

Wallmart has leveraged bigdata and analytics for a number of years now. The amount of data they collect is already at the level of petabytes and they save information about all of their customer transactions, more than 1 million per hour, and it is mainly related to receipts and purchases.  Even with a single source of data, it already provides a lot of insights when properly stored and analyzed. Today, the average retailer in the US has about 700 Terabytes of stored data. This is huge already. Processing this data requires often big investments especially if you want to process the data in real time.  But bigdata is not only for big enterprises it can also be offered as a service and a number of service providers are moving into analytics as a service, this will revolutionize the way retailers do business.

Connecting the right type of data sources could provide value even for the smallest companies. Most retailers own information collected at the point of sale in terms of receipts, purchases done, but that is not the only valuable source. Using video surveillance cameras to analyze customer behavior or combining information from mobile devices or social media could be other valuable sources. Much of this information comes today from mobile and M2M devices, but that data is often stored in silos and it can be difficult to get actionable insights from this data.  I think the key for smaller sized retailers will be to expose their own data assets, get that data enriched and get analytics as a service from specialized service providers. The question is now, how can these insights be used and monetized?

Productivity improvements and cost savings

A first application of properly processing retail data is to achieve cost savings. Knowing which items are in demand ahead of time and knowing what will sell can help to optimize inventory and supply chains. Utilizing big data could increase retailers’ operating margins by up to 60%, this could include savings in marketing, merchandising as well as supply chain.

Looking ahead, I could easily imagine in a not so far future, that bigdata and analytics could also become a means to improve in store customer service. Imagine a clerk in a fashion store that would know as you enter his store what you like in terms of clothes, what you sizes are, what colors you prefer and understands your purchasing behavior, he could serve you faster and better. This would result in more customers served and a better sales ratio.

Generate revenues from customer Insights, sentiment and behavior analysis

Another application of bigdata is to improve customer loyalty and reduce churn. To continue with the Wallmart reference, they have developed a product that allows them to reach customers, or friends of customers, who have mentioned something online to inform them about that exact product and include a discount. Using social media and combining that with purchasing data and contact information it allows them to implement pro-active measures and hence avoid churn or even attract new customers. This creates loyalty and strenghtens the brand image.

Retaining customers has also to do with the experience they get when interacting with the retailer, this is in part achieved by delivering relevant ads, offers, and promotions. Personalizing offers means making offers more revelant to a customer by having specific knowledge of that customer’s profile.

Securing increased sales is also related to the ability to rapidly adjust to the competitive environment. Most of us are sensitve to pricing and customers will often look for peer reviews to find the best place and time to buy an item. Being able to quickly adjust to this competitive environment will be key for increasing sales and sometimes even for survival.

M2M and mobile devices will play an increasing role in customer profiling.

There are today many data sources that are already in digital format and could be used in a big data context. From a retailer context we have customer information, this is often tied to the purchases done in that store through an opt-in loyalty program. Then there are the product catalogues and the actual status of the inventory.

The next step in personalization is coming from Mobile Devices, through applications which will let customers specify preferences, provide navigation capabilities to be guided to the right store, get relevant advertising and also get recommendations from other customers on the items they are interested in.

I think though that the biggest disruption might come from a smart usage of video cameras (e.g. surveillance camera) and other M2M sensors. Images from cameras in a store could help identifying who you are and support retailers to pul up a full profile of your customer preferences. Most of us tend to trade in privacy for convenience, so I have not doubt that every individual will sooner or later have pictures of himself tagged somewhere on the net and that could be the starting point to pull more information about you as a customer. But there are other mechanisms to identify you as a person as for instance foursquare check-ins or the social media profile you are using when logging to wifi.

Location information is going to be quite important to target shoppers. This could be obtained from mobile systems, wifi or tracking from video cameras. Video cameras could also be used to analyze how customers are focusing on different items in the store. This could help understand interest for certain items and maybe also decide when and what to discount. Another application of video camera is for instance to get the right sizes of the customers when entering a fashion store.

Digital signage in shopping malls and retail store can also be used to provide tailored messages and personalized ads to shoppers in or outside the store. A few seconds is probably enough to reach out to a shopper with the right message and the impact of the advertisement can probably be measured if the sale is done within the next  hour. Digital signage could also be used with cameras to analyze shopping carts and make a last second recommendation on something you might have forgotten to buy.

Other smart devices collecting data about users could be smart mirrors,  that record reactions when trying out clothes or screens that provide a personalized greeting when entering a store.

The future of retail – Smart Shopping malls

If I picture the future of retail I would imagine a smart shopping mall where customer experience is at the heart of this business. I would enter the mall and get greeted with a personalized message, the map of the mall would be downloaded to my mobile and I would get a recommendations on a few stores to visit. I would get discounts for the items I been looking for, especially the I have been reading recommendations about on the internet. This would be part of the profile information the mall is pulling up for me. As I walk and watch the digital signs I see more information about the products I am interested in. I enter a shop. The clerk again greets me and guides me directly to the item I am looking for and provides also alternative color options based on my preferences. Not much to ask me since he has all this information on a tablet.  As the clerk knows I am price sensitive he gives me an extra discount. If I picture the future of retail I would imagine a smart shopping mall where customer experience is at the heart of this business and where bigdata and analytics provide value at every step .

Why Telco’s should invest in bigdata capabilities

new context

What is bigdata and what are key operator data assets?

In a nutshell bigdata is about getting actionable insights from data which is available in huge volumes, has high velocity and is often found in a big variety. These are the so called 3V’s of big data. While the volume definition is very clear, with velocity we mean data which is of value only for a short period of time, possibly a few minutes or less. The variety of the data is also a bit difficult to understand at first. Here we talk primarily about the huge amount of unstructured data which is today hardly harvested, as an example you can think of insights found in text, images and video.

Data amounts available to analyze are huge and grow exponentially. An IBM study showed that 90% of all data available in the world has been produced over the last two years only. Operators sit today on huge amounts of data themselves, but few have started to monetize that knowledge neither internally nor externally.

There is probably different ways to segment the operator’s data assets but the taxonomy I used below is simple enough so that anybody can get a feeling for the types of use-cases that could be monetized by Telco’s. This taxonomy is based on different types of contexts which provide value to the telco or to a third party.

  • Operators have today a good understanding of the situational context of their users (i.e. position information, whether I am in mall, in my car or in an movie theater),
  • they understand my behavioral context through my devices I am using (am I in a call, watching my IPTV or browsing the web and which sites)
  •  and they know about my social context (most frequent people I call, my preferences through the content I am consuming, the places I go, which social networks I belong to and of course my billing information)

Now for privacy reasons much of this data cannot be exposed without the consent of the end-user,  unless there is an opt-in close when subscribing to a service. In any case the data can most of the time be made anonymous and brokered as such.

The opportunities for Telco’s are numerous but can be seen as internal and external opportunities. Improving retention and revenues by becoming more customer centric and better managing the customer experience is an internal usage of these insights, while an external usage is to provide new services to enterprises that will help them to improve their own efficiency and decision processes. In this post I will only cover the external bigdata opportunities.

Targeting enterprises with BigData Services and strengthening cloud offerings

From a 2102 Gartner study you can see that most industries are in the process of investing or a planning to invest in Bigdata technology of the next 2 years. Rather than further increasing ICT spending and OPEX in the enterprise their is an opportunity for operators to provide cloud services and become the technology partner for bigdata.

Gartner-diagram

While telcos are probably a bit late to the market with cloud services compared to companies like Amazon, operators have still a privileged position when it comes to trust, security and service level expectations. Telco’s are still privileged partner for larger enterprises, industry verticals and government mainly because of their local presence. This is why they need to invest in this area now or necessity will take over to find alternative providers.  For Telco’s the opportunities in bigdata when part of a cloud offering, can probably count the folowing non exhaustive list.

Storage provider – Data-as-a-service

For enterprises that collect or want to collect data, they will rapidly have to consider how to store that data in a cost efficient and scalable manner. This is where telco’s data center could be used to provide elasticity and handle that demand in a cost efficiency way. Of course depending on the type of enterprise data, there are important questions to address when it comes to privacy, integrity and security of the data, but operators are probably well positioned to offer Data-as-a-Service to different industry vertical according to the enterprises expectations.

Analytics-as-a-Service Provider

The tools and processes needed to gain insights require major changes and investments in the enterprise. Now even if you have data and can store most of your data, it sill requires new tools and competences to be able to gain insights from that data. Telco’s  could play a major role in taking bigdata to the masses, especially toward the SME’s where efficiency and agility is so important. Providing business intelligence reports, visualization capabilities as well as dashboard views are few of the Basic mechanisms to put in place.

Data Enrichment and brokering – Enriched Insights provider

The data operators own has even more value if consolidated or federated with other data sources like social media, government or enterprise data. If this data can be exposed and analyzed it will provide further insights that a government, large enterprise or industry verticals will be ready to pay for.

Become a strategic partner for enterprises, industry verticals, governments and the public sector

A said earlier, operators have a privileged position when interfacing governments and regulators and should leverage that position. In order to stay competitive and sustain economic growth governments need to take faster and smarter decisions. This often means collecting and analyzing more data to improve the decisioning process. By leveraging on bigdata technology, federation and analytics, operators could find new means to address value chains for which they were today only providing connectivity and communications.

I won’t be able to cover all bigdata opportunities where telco’s could provide value, but nelow you will find a couple of examples where big data could be useful and it will show how that could help operators position themselves in new value chains.

M2M and BigData

A simple example that could be considered here is how an operator can use big data to combine multiple structured and unstructured data sources to provide additional value to vehicle fleet management solution.  Analyzing social media streams, outgoing calls or lack of movement along certain roads, can help provide real-time information that could be useful for redirecting vehicles or a whole fleet. In addition the operator can combine this services with automated communication to guide a track, bus or taxi toward the right destination in real-time. Automative industry, Health care, Telemetry and most industries using M2M sensors could benefit from bigdata and analytics to increase the value or services offered by operators.

Retail and Advertising

For advertisers and retailers targeting the right audience and the ability to prove that you can address the right audience is worth heavy dollars. The propensity to buy for a customer is heavily correlated to how much know about that person, it tastes and preferences. Operators have a lot of information about the user profile, but more could be collected through the various portals and even through packet inspection. By analyzing consumer likes on social networks, purchases on amazon or  zapping patterns on an IPTV system a lot can be learned and shared.

Urban Planning

Urban planning can certainly benefit from the insights operators have in terms of road traffic development and customer segmentation in order to improve build-out of new infrastructure, including roads, housing, schools and hospitals.

Security Services

Take the example of a mall where you have cameras monitoring everything that happens. Quickly analyzing what happens over multiple video streams to identify a theft and identify the person would require real-time processing and federation over multiple data sources. Operators could be a natural broker for this type of use-case.

… and many more industries can be addressed

Many more industries and verticals could benefit from bigdata investments, but it would be difficult top properly cover all of this in one post. If there is interest from the community I will write separate post on how big data can help Education and Healthcare, which are two topics that deserve their own analysis.

Finally, BigData as a means for Telco’s to become a community as well as a lifestyle provider?

For telco’s there is potentially more at stake. Bigdata provides opportunities for operators to become more customer centric and become more relevant in their consumer’s lives. If done right they could become this lifestyle provider they have always dreamt off becoming.  By strengthening relationships with governments and selected enterprises they could also take a more active role in the community and by that increase their relevance both for consumers and enterprises.